def matToV4(m):
v= p4(0,0,0,0)
v.x=m[0][0]
v.y=m[1][0]
v.z=m[2][0]
v.w=m[3][0]
return v
def check_jobs():
p4i = p4.p4(port = config.p4_port,
client_executable = config.p4_client_executable,
user = config.p4_user,
password = config.p4_password)
jobs = p4i.run('jobs')
failures = 0
for j in jobs:
try:
p4i.run('job -o %s' % j['Job'])
except p4.error, msg:
# "Job '%s' doesn't match the jobspec:"
print str(catalog.msg(1008, j['Job']))
print str(msg)
failures = failures + 1
configure_name = string.lower(config.configure_name)
configure_module = __import__('configure_' + configure_name)
config = configure_module.configuration(config)
dt_name = 'dt_' + string.lower(config.dt_name)
dt_module = __import__(dt_name)
dt = getattr(dt_module, dt_name)(config)
# 4. MAKE A PERFORCE INTERFACE AND A "DEFECT TRACKER" FOR PERFORCE
p4_interface = p4.p4(client = ('p4dti-%s' % socket.gethostname()),
client_executable = config.p4_client_executable,
password = config.p4_password,
port = config.p4_port,
user = config.p4_user,
config_file = config.p4_config_file,
logger = config.logger)
# 5. MAKE THE REPLICATOR AND INITIALIZE IT
r = replicator.replicator(dt, p4_interface, config)
# A. REFERENCES
#
# [GDR 2000-09-13] "Replicator design"; Gareth Rees; Ravenbrook Limited;
# 2000-09-13;
from p1 import p1
from p2 import p2
from p3 import p3
from p4 import p4
import cv2
# Analyze two objects picture
image = p1('two_objects.pgm', 100)
cv2.imwrite('binary_image.pgm', image)
labeled_image, index_list = p2(image)
cv2.imwrite('labeled_image.pgm', labeled_image)
cv2.imshow('image', labeled_image)
cv2.waitKey(0)
info = p3(labeled_image)
database = info[0]
cv2.imwrite('output_image.pgm', info[1])
# Analyze many objects picture
image = p1('many_objects_1.pgm', 100)
labeled_image, index_list = p2(image)
output_image = p4(labeled_image, database)
cv2.imwrite("Analysis_Result.pgm", output_image)
img = cv2.imread("Analysis_Result.pgm", 0)
cv2.imshow('image', image)
cv2.waitKey(0)
## p3
result = p3(img)
p3_output = result[1].copy()
p3_output = img_scale(p3_output)
cv2.imwrite('p3_two_objects.jpg', p3_output)
## p4
query = cv2.imread('many_objects_1.pgm', cv2.IMREAD_UNCHANGED)
query = p1(query, 81)
query = cv2.copyMakeBorder(query, 1, 1, 1, 1, cv2.BORDER_CONSTANT, 0)
query = p2(query)
query = query[1 : query.shape[0] - 1, 1 : query.shape[1] - 1]
p4_output = p4(query, result[0])
p4_output = img_scale(p4_output)
cv2.imwrite('p4_many_objects_1.jpg', p4_output)
query = cv2.imread('many_objects_2.pgm', cv2.IMREAD_UNCHANGED)
query = p1(query, 81)
query = cv2.copyMakeBorder(query, 1, 1, 1, 1, cv2.BORDER_CONSTANT, 0)
query = p2(query)
query = query[1 : query.shape[0] - 1, 1 : query.shape[1] - 1]
p4_output = p4(query, result[0])
p4_output = img_scale(p4_output)
cv2.imwrite('p4_many_objects_2.jpg', p4_output)
# Run sequential labeling algorithm
labeled_img = p2.p2(binary_img)
# Save labeled image
cv2.imwrite("img_labels.pgm", labeled_img)
# Print object attributes
obj_info_arr = p3.p3(labeled_img)
obj_database = obj_info_arr[0]
for i in range(len(obj_database)):
print "Object # " + str(i + 1) + ":\n"
print "label = " + str(obj_database[i]['object_label'])
print "x position = " + str(obj_database[i]['x_position'])
print "y position = " + str(obj_database[i]['y_position'])
print "minimum moment = " + str(obj_database[i]['min_moment'])
print "orientation = " + str(obj_database[i]['orientation'])
print "roundness = " + str(obj_database[i]['roundness'])
print "\n"
# Save overlay image
cv2.imwrite("img_overlay.pgm", obj_info_arr[1])
# Check if similar two objects can be detected in 'many objects' image
gray_in = cv2.imread("many_objects_1.pgm", cv2.CV_LOAD_IMAGE_GRAYSCALE)
binary_img = p1.p1(gray_in, gray_level_threshold)
labeled_img = p2.p2(binary_img)
overlays_out = p4.p4(labeled_img, obj_database)
# Save recognized objects image
cv2.imwrite("img_similar_objects.pgm", overlays_out)
import numpy as np
import cv2
from p1 import p1
from p2 import p2
from p3 import p3
from p4 import p4
binary_img = p1('two_objects.pgm', 135)
labels_in = p2(binary_img)
database_in = p3(labels_in)
labels_many = p2(p1('many_objects_1.pgm', 135))
p4(labels_many, database_in)
plt.imshow(img_two_objects_binary * 255, 'gray')
plt.show()
# task 1b
img_two_objects_labeled = p2(img_two_objects_binary)
cv2.imwrite('two_objects_label.pgm', img_two_objects_labeled)
plt.imshow(img_two_objects_labeled * 30, 'gray')
plt.show()
# task 1c
database, img_two_objects_overlay = p3(img_two_objects_labeled)
print database
plt.imshow(img_two_objects_overlay)
plt.show()
# task 1d
img_many_objects_1 = cv2.imread('many_objects_1.pgm', 0)
img_many_objects_2 = cv2.imread('many_objects_2.pgm', 0)
img_many_objects_1_binary = p1(img_many_objects_1, 127)
img_many_objects_2_binary = p1(img_many_objects_2, 127)
img_many_objects_1_labeled = p2(img_many_objects_1_binary)
img_many_objects_2_labeled = p2(img_many_objects_2_binary)
img_many_objects_1_overlay = p4(img_many_objects_1_labeled, database)
img_many_objects_2_overlay = p4(img_many_objects_2_labeled, database)
plt.imshow(img_many_objects_1_overlay)
plt.show()
plt.imshow(img_many_objects_2_overlay)
plt.show()
# This is the end of task 1
